Birdsong is a learned, complex behavior that is organized on multiple timescales that are analogous with aspects of human speech such as prosody and phonemes. It has been proposed that songs are encoded over neural circuits on a single fine timescale only;however, these studies have been restricted to feed forward control in the forebrain. The general aim of this research is to examine how recurrent brainstem-forebrain circuitry (BFC) represents the rhythmic aspects zebra finch song on broader scales. I hypothesize that BFC integrates corollary discharge of forebrain activity with the generation of commands for upcoming vocalizations. I will test this hypothesis with chronic, single-unit recordings in two key areas of BFC during adult song production. First, I will record from the dorsal region of forebrain nucleus RA, which is currently understood to encode the respiratory activity which accompanies song. Second, I will record from brainstem nuclei PAm and DM, both of which receive inputs from dorsal RA and project back towards the forebrain via the thalamus;currently very little is known about how song is encoded over these subpopulations. I will compare the timing of spike trains in each of these areas with syllable onsets, offsets and lengths. Importantly, I will exploit trial-to-trial variability in song timing and sequencing as a critical degree of freedom in this analysis: although zebra finch song is highly stereotyped, it also exhibits subtle but systematic variability in timing and sequencing that may be directly related to the motor code. Naturally occurring perturbations such as these could reveal especially weak links in neural networks that are more easily disrupted under more severe conditions such as a disease. Analogous timescales in human speech have been identified and are implicated in aphasias such as stuttering and lesions to key areas such as the cerebellum. Thus, the discovery of how "song prosody" is encoded in zebra finch song may reveal important insights into the physiological basis for human speech and known pathologies. Public statement: Birdsong has provided a exciting opportunity to investigate how the brain produces complex, vocal communication. Songs have a temporal organization that is very similar to human speech, but so far the neural basis of this organization remains poorly understood. The current study is aimed at how neural circuits encode the rhythm and order of individual vocalizations in song. This research may thus reveal important insights into disorders that affect similar elements of human speech, such as stuttering.